Chapter

Chapter 2. Reinvigorating Growth in the Caribbean

Author(s):
Krishna Srinivasan, Inci Otker, Uma Ramakrishnan, and Trevor Alleyne
Published Date:
November 2017
Share
  • ShareShare
Show Summary Details
Author(s)
Marcos Chamon, Joshua Charap, Qiaoe Chen, Daniel Leigh, Franz Loyola and Lulu Shui 

Introduction

Growth in the Caribbean has been disappointing in recent decades. Averaging 2.1 percent per year since 2000, real GDP growth for the Caribbean has been half that of other emerging market and developing economies (EMDEs) and two-thirds that of non-Caribbean small states (Figure 2.1 and Table 2.1).1 The growth weakness is concentrated among tourism-intensive Caribbean economies, which grew annually by only 1.6 percent (0.8 percent in per capita terms). Such low rates of growth complicate job creation, the raising of Caribbean wages and social conditions toward advanced economy levels, and management of the region’s significant burden of public and private sector debt. Commodity exporters have seen faster growth during this period (3.7 percent per year), largely reflecting positive effects from the commodity price boom in the 2000s. More recently, however, growth in the commodity-exporting countries has also slowed or turned negative.

Figure 2.1.Real GDP, 2000–15

(Cumulative; index, 2000 = 100; country-group medians)

Sources: IMF, World Economic Outlook; Penn World Table; and IMF staff estimates.

Note: EMDE = emerging market and developing economy.

Table 2.1.Economic Growth, 2000–15(Percent per year; median)
Real GDP Growth
CountryTotalPer Capita
Caribbean2.11.2
Tourism-Intensive Caribbean Economies1.60.8
Antigua and Barbuda1.60.8
The Bahamas0.6−0.4
Barbados0.80.5
Dominica1.72.1
Grenada2.21.6
Jamaica0.60.1
St. Kitts and Nevis2.21.2
St. Lucia1.30.6
St. Vincent and the Grenadines2.12.0
Commodity-Exporting Caribbean Economies3.73.0
Belize3.40.9
Guyana3.12.9
Suriname4.23.0
Trinidad and Tobago3.93.6
Memorandum
Non-Caribbean EMDEs4.22.6
Non-Caribbean Small States3.01.4
Non-Caribbean Commodity Exporters4.22.4
Advanced Economies1.91.1
Sources: IMF, World Economic Outlook; Penn World Table; and IMF staff estimates.Note: EMDE = emerging market and developing economy.
Sources: IMF, World Economic Outlook; Penn World Table; and IMF staff estimates.Note: EMDE = emerging market and developing economy.

This chapter provides an overview of the conditions that have limited Caribbean growth by synthesizing insights from existing research and updating estimates on drivers of growth. It begins by comparing growth of Caribbean economies with that of peer groups, including countries at a similar level of development. To shed light on whether sluggish growth reflects slow accumulation of capital and labor or weak productivity growth, the chapter conducts a growth accounting exercise. Next, to highlight policy priorities for raising growth, the analysis ranks Caribbean economies and their peers according to 20 country characteristics relevant for growth. The chapter then provides estimates of the extent to which disappointing growth in tourism-intensive Caribbean economies can be explained by identified areas of weakness using growth equations estimated with the latest cross-country data. Finally, based on these growth equations, the chapter estimates the medium-term increase in growth that could come from addressing structural challenges, then discusses necessary policy reforms.

Caribbean Growth in Global Perspective

Since 2000, economic growth in all tourism-intensive Caribbean economies has been slower than for most EMDEs and non-Caribbean small states (Table 2.1). It has, on average, been even lower than for advanced economies, based on both overall GDP and on income per capita (converted at purchasing-power-parity exchange rates that more accurately reflect differences in the cost of living across countries). While the global financial crisis surely affected Caribbean economies severely, it is not an adequate explanation for their relatively weak average growth over this decade and a half. The shock of the global crisis may have amplified underlying domestic vulnerabilities. Commodity-exporting Caribbean economies have grown about as fast as non-Caribbean commodity exporters during this period—slightly slower in total GDP terms and slightly faster in per capita terms.

One may argue that the slow growth of Caribbean economies reflects the fact that they have reached a higher level of per capita income than most EMDEs, and therefore have less scope for rapid growth. This simple proposition is, however, not entirely valid for the Caribbean economies. The per capita income of Caribbean countries is well above the EMDE average, with a median level of US$10,600 in purchasing-power-parity terms as of the end of 2015, compared with the EMDE median of US$6,800. Standard growth theory predicts that these countries would grow more slowly because there is less of a gap with advanced economies to bridge through the accumulation of capital and technological leapfrogging. The experience of 179 economies since 2000 is consistent with the idea that countries at a higher income level grow more slowly (Figure 2.2).

Figure 2.2.Growth per Capita versus Initial per Capita Income, 2000–15

Sources: IMF, World Economic Outlook; Penn World Table; and IMF staff estimates.

Note: For tourism-intensive Caribbean economies, size of circles reflects population size. Data labels in figure use International Organization for Standardization (ISO) country codes.

Growth in tourism-intensive Caribbean economies has, however, been less than could have been expected on the basis of their per capita incomes. These countries should have grown at about 2.3 percent per year in per capita terms, as compared to the registered growth that has averaged 0.8 percent per year (Figure 2.3).2 This puzzling weakness does not apply to commodity-exporting Caribbean economies, which grew slightly faster than could have been expected during this period, reflecting the commodity boom of the 2000s.

Figure 2.3.Actual and Predicted Average GDP Growth per Capita, 2000–15

(Percent per year)

Sources: IMF, World Economic Outlook; Penn World Table; and IMF staff estimates.

Note: “Actual” denotes mean of sample available for estimating growth equation.

A defining feature of Caribbean growth compared with that of other regions has been the weak contribution of total factor productivity (TFP), which measures the overall productivity of both labor and capital, and reflects such elements as technology. A simple growth decomposition exercise suggests that TFP growth since 2000 has been near zero or negative, unlike for non-Caribbean small states and other EMDEs (Figure 2.4).3 On average, during 2000–15, TFP growth contributed –1.4 percentage points to annual growth for tourism-intensive Caribbean economies and 0.6 percentage points for commodity-exporting Caribbean economies, compared with larger positive contributions for non-Caribbean small states and EMDEs. This finding is consistent with other studies that point to falling or stagnating productivity in the Caribbean (for example, Ruprah, Melgarejo, and Sierra 2014; Thacker and others 2013) and with a recent study of productivity and innovation at the level of small and medium-sized Caribbean enterprises (Dohnert, Crespi, and Maffioli 2017).

Figure 2.4.Contributions to Average GDP Growth per Capita, 2000–15

(Percent per year)

Source: Authors’ calculations.

Note: Figure reports results for commodity-exporting Caribbean economies and tourism-intensive economies, non-Caribbean small states, and non-Caribbean EMDEs. EMDE = emerging market and developing economy.

What’s the Holdup?

To shed light on which country characteristics and policies may explain the slow growth of tourism-intensive Caribbean economies, this section proceeds in two steps. First, it compares performance with EMDE peer groups along dimensions that are relevant for growth and prosperity.4 The second step uses estimated growth equations to quantify the extent to which growth underperformance could be attributed to the structural weaknesses identified in the first step.

Diagnostic of Characteristics Relevant for Growth and Prosperity

A review of 20 country characteristics relevant for medium-term growth and prosperity reveals areas in which Caribbean economies rank poorly compared with other EMDE country groups (Table 2.2). For each of the 20 indicators, the color of each cell indicates the relative ranking of a country group compared with the others. Red indicates a less favorable ranking and green indicates a more favorable ranking. For example, the greater the ease of doing business compared with other country groups, the more green the cell color, and the higher the violent crime rate, the more red the cell color.

Table 2.2.Heat Map: Characteristics Relevant for Growth and Prosperity(Ranking across country groups)
Tourism-Intensive CRB EconomiesCommodity-Exporting CRB EconomiesCentral AmericaSouth AmericaCEEEM AsiaLICsNon-CRB Small States
Violent Crime
Emigration
Skilled Emigration
Human Capital
Natural Disasters
Government Debt
Taxation
NPLs
Ease of Doing Business
Competitiveness
Bureaucracy
Regulatory Quality
Rule of Law
Control of Corruption
Trade Openness
Trade Tariff
Trade Connectivity
GVC Participation
High-Tech Export Share
Export Diversification
Sources: Barro-Lee 2013 data set; IMF, Diversification database; IMF, World Economic Outlook; Institute for Employment Research, brain-drain data set; Penn World Table; World Bank, Ease of Doing Business Index; World Bank, World Development Indicators; World Bank, World Governance Indicators; and World Economic Forum, Global Competitiveness Index.Note: See Annex Table 2.1.1 for data descriptions. Red = less favorable ranking; green = more favorable ranking. Table based on 2015 or most recent data. For natural disaster damage, table based on historical average (2000–15). CEE = Central and Eastern Europe; CRB = Caribbean; EM Asia = emerging market economies in Asia; GVC = global value chain; LICs = lower-income countries; NPLs = nonperforming loans.
Sources: Barro-Lee 2013 data set; IMF, Diversification database; IMF, World Economic Outlook; Institute for Employment Research, brain-drain data set; Penn World Table; World Bank, Ease of Doing Business Index; World Bank, World Development Indicators; World Bank, World Governance Indicators; and World Economic Forum, Global Competitiveness Index.Note: See Annex Table 2.1.1 for data descriptions. Red = less favorable ranking; green = more favorable ranking. Table based on 2015 or most recent data. For natural disaster damage, table based on historical average (2000–15). CEE = Central and Eastern Europe; CRB = Caribbean; EM Asia = emerging market economies in Asia; GVC = global value chain; LICs = lower-income countries; NPLs = nonperforming loans.

Notable structural and macroeconomic impediments to growth include the following:

  • Brain drain. Caribbean countries stand out from their peers regarding outward migration of skilled human resources. The share of nationals with tertiary education (a proxy for high skills and knowledge) living abroad is about 76 percent, several times the rate for other EMDE regions, and 50 percent greater than the rate for non-Caribbean small states (Figure 2.5). Skilled emigration lowers growth by impairing the economy’s stock of skills and knowledge (human capital).5IMF (2017b) finds that the exodus of skilled labor has contributed to rising wage costs, with negative effects on external competitiveness. Remittances from Caribbean emigrants have been sizable, supporting investment and education and fostering commercial linkages, but IMF (2017b) concludes that the overall impact of emigration on growth has been negative for Caribbean countries.
  • Natural disasters. Caribbean countries are highly vulnerable to natural disasters because many of them are located in the cyclone and hurricane belts bordering the equator. Natural disasters reduce economic output by destroying crops, infrastructure, and, as Chapter 3 highlights, reducing tourism arrivals. Average annual physical damage from natural disasters amounted to 2.3 percent of GDP for tourism-intensive Caribbean economies since 2000, and 1.3 percent of GDP for commodity-exporting Caribbean economies, more than in non-Caribbean small states. Climate change is likely to exacerbate these challenges through rising sea levels (IMF 2016b).
  • Crime. Caribbean economies have some of the highest violent crime rates in the world, with an average homicide rate of about 24 per 100,000 inhabitants in 2015, more than three times the rate in non-Caribbean small states.6 Crime results in significant social costs, which tend to be concentrated among the most vulnerable members of society. As a detailed study by the staff of the Inter-American Development Bank explains, crime impedes economic growth by discouraging tourism and business investment, reducing worker productivity, and diverting government spending from investment in health, education, and productive infrastructure (Jaitman 2017). Jaitman (2017) estimates that direct crime-related costs amount to about 3.6 percent of GDP for the Caribbean.7 Violent crime is a particularly acute challenge for tourism-dependent economies because it depresses both demand for tourism and the supply side of the economy.
  • Government debt burden. Tourism-intensive Caribbean economies have accumulated the highest levels of government debt among EMDEs (Chapter III.1). Commodity-exporting economies have had lower levels of debt, but they recently saw a rapid increase. High debt presents costs and risks that weigh on medium-term growth, including distortions from higher taxes needed to service the debt, displacement of productive government spending, and concerns associated with the possibility of default, which feeds uncertainty about future taxation and inflation. A number of studies have found a negative association between high government debt and growth for Caribbean economies, including, most recently, Thacker and others (2013).
  • Financial sector strains. The banking sector in a number of Caribbean economies is burdened by high levels of nonperforming loans (NPLs), averaging 12 percent of the value of loans in tourism-intensive Caribbean economies, and 9 percent in commodity-exporting Caribbean economies, well above the levels for other EMDE groups. NPLs have a negative impact on bank profitability, increase bank vulnerability to shocks, and, in the absence of well-developed capital markets, constrain credit supply to finance productive investment and growth (Chapter IV.3). During 1995–2015, Caribbean economies underwent a number of episodes of financial stress, as documented by Laeven and Valencia (2013), which can have long-lasting effects on economic activity.
  • Business environment and competitiveness. Stronger protection of property rights and better, simpler regulations for businesses are important for encouraging private investment and durable growth. The challenges in this area are greater for the commodity-exporting Caribbean economies than for the tourism-intensive ones, as is illustrated by their relatively poor rankings on the World Bank Ease of Doing Business Index, the Global Competitiveness Index of the World Economic Forum, and the World Bank Worldwide Governance Indicators (perceptions of corruption, law and order, regulatory quality, and bureaucratic efficiency).8
  • Trade integration. Stronger trade links tend to raise long-term growth; the channels include enhanced competition and learning and productivity spillovers. The small size of Caribbean economies and the erosion of trade preferences extended to them (Cashin, Haines, and Mlachila 2010) accentuate the need for trade and market integration to overcome their constrained set of inputs, including the number of entrepreneurs, that can move to new and more productive activities as part of structural transformation and diversification (Hausmann and Klinger 2009). Caribbean trade openness—exports and imports as a share of GDP—is broadly comparable with other regions, but lower than could be expected based on the small size of Caribbean economies, which increases reliance on imports (Figure 2.6). Caribbean connectivity with the world trade network, as measured by the number of trading partners and their network centrality, as well as participation in global value chains (networks of production stages of goods and services across borders) remains limited compared with other regions.9 These limitations in part reflect the small size of Caribbean economies, but do constrain opportunities for diversifying and upgrading the complexity and technological content of exports.

Figure 2.5.Emigration Rates, 2010

(Share of country nationals living abroad)

Sources: Institute for Employment Research, brain-drain data set; and IMF staff estimates.

Note: High-skilled emigration relates to nationals with tertiary education. Data labels in figure use International Organization for Standardization (ISO) country codes.

Figure 2.6.Trade Integration and Country Size

Sources: IMF, Direction of Trade Statistics; and IMF staff estimates.

Note: Trade openness defined as sum of exports and imports as a share of GDP. Trade connectivity measured by five indicators from IMF (2017a) covering number of countries traded with as well as their centrality in the world trade network, combined in a composite index. CEE = Central and Eastern Europe; CRB = Caribbean; EM Asia = emerging market economies in Asia; LA = Latin America; LICs = low-income countries; MENA = Middle East and North Africa.

Each of these factors matters individually, but their cumulative impact is compounded, with the potential to cause persistently weak growth. For example, violent crime may cause emigration and brain drain, which in turn can make it harder to cope with high debt burdens because fewer skilled workers remain to shoulder the debt service. By the same token, tackling one obstacle to growth can facilitate responding to others.

Insights from Growth Equations

How much of the puzzlingly weak growth of tourism-intensive Caribbean economies can be explained by their structural challenges? Estimating medium-term growth equations using data for many countries can shed light on this question. The equations are estimated in the form of five-year non-overlapping intervals covering 1995–2015 for an unbalanced panel of up to 148 advanced and emerging market and developing economies, including the Caribbean economies under study.10 Specifications for which data are limited are restricted to 2010–15. The estimated equations include a full set of time fixed effects to take account of global shocks such as shifts in commodity prices or the global business cycle. To address concerns about reverse causality, the values for the right-hand side variables are for the year preceding each five-year growth interval. Since natural disasters are independent of growth, they are included contemporaneously in the estimated equations. To guard against the undue influence of outliers, the estimation results are reported based on Cook’s distance approach, as well as ordinary least squares.11 See Table 2.3.

Table 2.3.Per Capita Real GDP Growth Equation Estimates
1995–20152010–15
OLSCook’s DOLSCook’s D
(1)(2)(3)(4)
Log of Initial Income per Capita−1.267***−1.169***−1.690***−1.771***
Trade Integration0.366*0.492***0.435**0.525***
Human Capital2.176***2.479***−0.2801.373**
Natural Disaster−6.587**−10.069**−10.326−27.977**
Crime−34.831***−25.274***−30.900**−27.352***
Government Debt−0.487−0.683**−1.058**−1.303***
Financial Crisis−1.029***−1.037***
NPLs−1.351−5.909*
Ease of Doing Business6.464*5.715***
Dummy, 2001–05−0.178−0.468
Dummy, 2006–10−0.847***−1.003***
Dummy, 2011–15−1.230***−1.349***
Constant13.470***12.170***14.544***14.312***
Observations439419127118
R20.1930.2550.2970.418
Source: Authors’ estimates.Note: Outliers excluded using Cook’s distance (D) method. Sample includes up to 148 advanced and emerging market economies (unbalanced panel). NPLs = nonperforming loans; OLS = ordinary least squares. *, **, and *** denote statistical significance at the 10, 5, and 1 percent level, respectively.
Source: Authors’ estimates.Note: Outliers excluded using Cook’s distance (D) method. Sample includes up to 148 advanced and emerging market economies (unbalanced panel). NPLs = nonperforming loans; OLS = ordinary least squares. *, **, and *** denote statistical significance at the 10, 5, and 1 percent level, respectively.

The estimation results for growth drivers that are beyond the control of policymakers are intuitive. A higher initial level of per capita income is associated with slower growth, and the coefficient estimate is consistent with the literature. In addition, a number of external factors beyond the control of policymakers, captured by the equations’ time fixed effects, had significant effects on growth. The global financial crisis of 2008–09 and its aftermath and the EMDE slowdown in more recent years are examples of adverse global factors that may explain the respective negative time fixed effect coefficient estimates.

Structural characteristics for which Caribbean economies rank unfavorably have the expected association with growth. Higher violent crime is associated with slower subsequent medium-term growth.12 A higher level of human capital—measured based on the adjusted years of schooling of the population—is associated with faster growth. Natural disasters reduce growth over the five-year periods considered. A higher initial stock of government debt is associated with lower subsequent growth. Financial strains, whether measured by the occurrence of a banking, currency, or sovereign debt crisis (columns (1) and (2) in Table 2.3) or by the initial NPL ratio (columns (3) and (4)) are associated with weaker subsequent growth. A more favorable business climate is associated with stronger growth, as is greater trade integration, measured by the ratio of trade to GDP.

The estimation results suggest that domestic structural challenges explain the bulk of the growth underperformance of tourism-intensive Caribbean economies (Figure 2.7). For 2000–15, the estimates based on the models reported in Table 2.3 explain more than 75 percent of the shortfall of actual growth below the rate predicted based on the level of income per capita alone. For the 2010–15 subsample, the models explain more than 95 percent of the shortfall.

Figure 2.7.Explaining the Growth Underperformance of Tourism-Intensive Caribbean Economies

(Average growth in real GDP per capita; percent per year)

Sources: IMF, World Economic Outlook; Penn World Table; and IMF staff estimates.

Note: Estimates based on coefficients in Table 2.3 (columns (2) and (4), for 2000–15 and 2010–15, respectively).

Policies for Reinvigorating Growth

This chapter’s analysis suggests ample scope for raising growth in the Caribbean by tackling the region’s structural challenges. For illustrative purposes, the analysis simulates the increase in growth that would come from an improvement in the policy-related factors to the top decile of small states, based on the growth equation estimates (Table 2.4). The assumed adjustments should not be interpreted as overnight outcomes because they could take years or even decades to achieve.13 Some of the reforms, such as government spending cuts or tax hikes aimed at reducing the stock of public debt, may even cause short-term negative effects on growth. For crime and disaster damage limitation, the simulations assume a reduction by one-half from the current level, and for trade integration, they assume an increase to the top quartile of small states. These are still ambitious goals, though less dramatic than improving to the top fifth percentile of small states. For government debt, a reduction to 60 percent of GDP is assumed.

Table 2.4.Illustrative Medium-Term Growth Gains(Caribbean median)
Baseline Level1Illustrative Level (top decile of small states)2Medium-Term GDP Growth Gain (percentage points)3
NPLs (percent of loans)10.942.330.51
Ease of Doing Business (index)57.5065.840.48
Disaster Damage (annual, percent of GDP)2.241.120.33
Crime (homicide rate per 100,000)20.9810.490.29
Government Debt (percent of GDP)80.6560.000.28
Human Capital (index)2.322.530.19
Trade (percent of GDP)93.02122.750.16
Total2.23
Source: IMF staff estimates.Note: NPLs = nonperforming loans.

2015 level except for disaster damage (historical average). Trade integration based on trade-to-GDP ratio. See Annex Table 2.1.1 for data descriptions.

For crime and disaster damage, reduction by 50 percent is assumed. For government debt, reduction to 60 percent of GDP is assumed. For trade, a rise to the top quartile of non-Caribbean small states is assumed.

Estimates based on coefficients in Table 2.3 (column (4)).

Source: IMF staff estimates.Note: NPLs = nonperforming loans.

2015 level except for disaster damage (historical average). Trade integration based on trade-to-GDP ratio. See Annex Table 2.1.1 for data descriptions.

For crime and disaster damage, reduction by 50 percent is assumed. For government debt, reduction to 60 percent of GDP is assumed. For trade, a rise to the top quartile of non-Caribbean small states is assumed.

Estimates based on coefficients in Table 2.3 (column (4)).

The estimated illustrative growth gains associated with these structural adjustments vary across countries, reflecting their different challenges, but are in the range of 1–3 percentage points per year. Such an increase in growth would represent significant progress toward increasing prosperity (Figure 2.8). The policy priorities differ by country. In Jamaica, for example, the largest growth gains come from reducing government debt and crime. For St. Lucia, the largest gain comes from addressing NPLs, and for Suriname, from improving the business environment. The gains from greater intra- and interregional trade integration reported here complement the more detailed estimates of recent IMF research on the effects of expanding trade networks and achieving greater participation in global value chains for economies in Latin American and the Caribbean (IMF 2017a).

Figure 2.8.Reinvigorating Growth: Illustrative Medium-Term Growth Gains

(Percentage points per year; deviation from baseline)

Source: IMF staff estimates.

Note: Estimates based on coefficients in Table 2.3 column (4). NPLs = nonperforming loans. Data labels in figure use International Organization for Standardization (ISO) country codes.

A set of self-reinforcing reforms, some of which Caribbean governments have already announced and are implementing, is needed to create an environment conducive to such medium-term gains. In conclusion, this chapter provides an overview of some of the main shared policy priorities for reinvigorating Caribbean growth.

  • Curbing and reversing brain drain. A comprehensive strategy is needed to create an environment that encourages people to stay and provides incentives for return migration. To retain highly skilled workers, who tend to leave for technological or scientific “hubs,” focusing on modernizing education and creating a critical mass of highly skilled workers is vital. Given the small size of Caribbean economies, achieving the required critical mass of skilled workers may require further promotion of labor market integration across the region.
  • Fighting crime. More work is required to assess the effects of existing public policies and crime prevention programs in the Caribbean, but, as set out in a comprehensive study by the staff of the Inter-American Development Bank (Jaitman 2017), a greater emphasis on prevention, alongside suppression, is warranted. Prevention elements should be integrated into social programs in health and education, especially among the youth population. Some promising initiatives are based on adapting programs that were successful in other jurisdictions. Trinidad and Tobago adapted Chicago’s CeaseFire program, which reduces violent crime through prevention and community-mobilization strategies, including intervention by “violence interrupters” who reach out to gang leaders and at-risk youth and mediate conflicts to reduce shootings and retaliatory violence (Ritter 2009). Jaitman 2017 also notes the effectiveness of moving toward restorative justice in Jamaica and Trinidad and Tobago, wherein victims and offenders mediate a restitution agreement. The criminal justice system should also be reformed to improve performance measurement and promote institutional accountability.
  • Deepening trade integration. The Caribbean Community is one of the region’s most important trade initiatives. Further progress toward the planned single market and economy in the Caribbean Community, which stalled in the aftermath of the global financial crisis, would lead to a further increase in trade. Caribbean economies would also be well served by greater integration with Latin American and the rest of the world. Interregional integration can be facilitated by new trade agreements, and there is scope in some cases to leverage current agreements more effectively, or deepen them to include areas beyond traditional trade issues. Deeper integration of Caribbean economies with global trade could also be achieved by reducing trade barriers, enhancing port infrastructure, fostering human capital formation, and supporting research and development (IMF 2017a).
  • Implementing sector-specific policies to support structural transformation. A number of studies argue in favor of sector-specific public intervention to promote structural transformation where markets have failed (for example, Hausmann and Rodrik 2006). For Caribbean economies, Hausmann and Klinger (2009) suggest that if small market size is preventing firms from moving to new and more productive activities, proactive government intervention in the form of “strategic bets” may be appropriate, informed by an inclusive public-private consultation process. Such intervention needs to leverage comparative advantage and fill gaps in public goods such as infrastructure, education, and regulatory issues, and avoid “picking winners” and subsidizing low-productivity sectors or firms through tax exemptions and firm-specific subsidies. Examples include supporting the setup of new tourism destinations, including for medical tourism, through advertising activities, training, nature conservation efforts, and the provision of transportation infrastructure; and supporting agricultural productivity by upgrading feeder roads, land-holding laws, and phytosanitary standards. Unlike larger economies, small states cannot place several bets in the hope that some will work and pay for those that do not.
  • Enhancing the ease of doing business. The policy priorities mentioned above would all help strengthen the ease of doing business in the Caribbean. The very small scale of most Caribbean economies makes removing barriers for businesses operating across these economies to achieve economies of scale particularly important. To promote financial development and access to credit, policy initiatives need to focus on institutional reforms, including strengthening property rights, enhancing the range of financial instruments, making credit information more available, and reducing the costs of financial intermediation.
Annex 2.1. Data Sources
Annex Table 2.1.1.Data Sources
IndicatorSource and Notes
Real GDPPWT, extended based on WEO
PopulationWEO
Capital StockPWT, extended based on WEO
LaborPWT, extended based on WEO
Violent CrimeWDI; homicides per 100,000 inhabitants
EmigrationIAB brain-drain data set; share of nationals living abroad
Skilled EmigrationIAB brain-drain data set; share of nationals with tertiary education living abroad
Human CapitalPWT, Barro-Lee (2013) data set; years of schooling scaled by assumed rate of return to education based on Mincer equation estimates (Psacharopoulos 1994)
DisastersEM-DAT; damage from natural disasters (percent of GDP)
Government DebtWEO; general government debt in percent of GDP
TaxationWDI; total revenue in percent of GDP
NPLsWDI, augmented by data collected for Chapter 12
Financial CrisisLaeven and Valencia 2013
Ease of Doing BusinessWorld Bank Ease of Doing Business Index distance to frontier (0–100)
CompetitivenessWorld Economic Forum Global Competitiveness Index
Bureaucratic EfficiencyWorld Bank WGI; index
Regulation QualityWorld Bank WGI; index
Rule of LawWorld Bank WGI; index
Control of CorruptionWorld Bank WGI; index
Trade OpennessWorld Bank WGI; index
Trade TariffWorld Bank WDI; average trade tariff (percent)
Trade ConnectivityIMF 2017b
GVC ParticipationIMF 2017b
High-Tech Export ShareWorld Bank WDI; percent
Export DiversificationIMF Diversification database; index
Note: Data are 2015 or latest available. For natural disaster damage, distribution for 2010–15 is reported.EM-DAT = International Disasters Database; GVC = global value chain; IAB = Institute for Employment Research (Brücker, Capuano, and Marfouk 2013); NPLs = nonperforming loans; PWT = Penn World Tables; WEO = IMF, World Economic Outlook; WDI = World Development Indicators; WGI = World Governance Indicators.
Note: Data are 2015 or latest available. For natural disaster damage, distribution for 2010–15 is reported.EM-DAT = International Disasters Database; GVC = global value chain; IAB = Institute for Employment Research (Brücker, Capuano, and Marfouk 2013); NPLs = nonperforming loans; PWT = Penn World Tables; WEO = IMF, World Economic Outlook; WDI = World Development Indicators; WGI = World Governance Indicators.
Annex 2.2. Growth Accounting Methodology

The conventional Cobb-Douglas output production function is used to calculate each factor’s contribution to growth. The analysis assumes that the production function includes capital and labor. The equation for output per capita is thus

where Y is output, K is the physical capital stock, L is employed labor, P is population, and A is total factor productivity (TFP). Differentiating with respect to time, and expressing the variables as growth rates yields the following:

The contribution of each factor is calculated as its growth rate multiplied by its share, with TFP as the residual. The analysis is based on the conventional assumption that the share of capital, α, is 0.35.

The physical capital stock is adjusted for the impact of natural disasters, which inflict significant damage on Caribbean countries. Without adjusting the capital stock, the contribution of capital would be overstated, and negative effects of natural disasters would be ascribed to TFP. Following Thacker and others (2013), the perpetual inventory method is used to compute the capital stock as follows:

where δt is the depreciation rate, It is investment, and NDt is natural disaster damage to the capital stock in year t.

Data for Y, K, L, P , and δ are taken from the Penn World Tables (9.0). For 2015 and 2016, the latest World Economic Outlook data are used to extend the series. The Emergency Events Database is the source of the disaster damage estimates, which are available in U.S. dollars.

References

    Barro, RobertJ., and Jong-WhaLee. 2013. “A New Data Set of Educational Attainment in the World, 1950-2010.” Journal of Development Economics104: 184–98.

    • Search Google Scholar
    • Export Citation

    Brücker, Herbert, StellaCapuano, and AbdeslamMarfouk. 2013. “Education, Gender and International Migration: Insights from a Panel-Dataset 1980-2010.” Institute for Employment Research, Nuremberg.

    • Search Google Scholar
    • Export Citation

    Cashin, Paul, ClearyHaines, and MontfortMlachila. 2010. “Caribbean Bananas: The Macroeconomic Impact of Trade Preference Erosion.” IMF Working Paper 10/59, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

    Cohen, Daniel, and LauraLeker. 2014. “Health and Education: Another Look with the Proper Data.” Unpublished, Paris School of Economics.

    • Search Google Scholar
    • Export Citation

    Cook, R. Dennis.1979. “Influential Observations in Linear Regression.” Journal of the American Statistical Association74 (365): 169–74.

    • Search Google Scholar
    • Export Citation

    Demirci, Ozge, EdwardMoreno, and JoyceChengWong. Forthcoming. “Does Crime Hinder Growth: An IV Approach to Latin America and the Caribbean.” IMF Working Paper, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

    Dohnert, Sylvia, GustavoCrespi, and AlessandroMaffioli. 2017. Exploring Firm-Level Innovation and Productivity in Developing Countries: The Perspective of Caribbean Small States. Washington, DC: Inter-American Development Bank.

    • Search Google Scholar
    • Export Citation

    Hausmann, Ricardo, and BaileyKlinger. 2009. “Policies for Achieving Structural Transformation in the Caribbean.” Private Sector Development Discussion Paper 2, Inter-American Development Bank, Washington, DC.

    • Search Google Scholar
    • Export Citation

    Hausmann, Ricardo, and DaniRodrik. 2006. “Doomed to Choose: Industrial Policy as Predicament.” Working Paper, Center for International Development, Harvard University, Cambridge, MA.

    • Search Google Scholar
    • Export Citation

    Hayashi, Fumio. 2000. Econometrics. Princeton, NJ: Princeton University Press.

    International Monetary Fund (IMF). 2016a. “Jamaica: Staff Report for the 2016 Article IV Consultation, Eleventh and Twelfth Reviews under the Extended Fund Facility for Jamaica.” IMF Country Report 16/181, Washington, DC.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund (IMF). 2016b. “Small State’s Resilience to Natural Disasters and Climate Change—Role for the IMF.” IMF Policy Paper, Washington, DC.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund (IMF). 2017a. “Cluster Report—Trade Integration in Latin America and the Caribbean.” IMF Country Report 17/66, Washington, DC.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund (IMF). 2017b. Regional Economic Outlook: Western Hemisphere. Washington, DC, April.

    Jaitman, Laura. 2017. The Cost of Crime and Violence: New Evidence and Insights in Latin America and the Caribbean. Washington, DC: Inter-American Development Bank.

    • Search Google Scholar
    • Export Citation

    Laeven, Luc, and FabianValencia. 2013. “Systemic Banking Crises Database.” IMF Economic Review61 (2): 225–70.

    Psacharopoulos, George. 1994. “Returns to Investment in Education: A Global Update.” World Development22 (9): 1325–43.

    Ritter, Nancy2009. “CeaseFire: A Public Health Approach to Reduce Shootings and Killings.” National Institute of Justice Journal. Issue No. 264 (November2009).

    • Search Google Scholar
    • Export Citation

    Ruprah, InderJ., KarlAlexander Melgarejo, and RicardoSierra. 2014. Is There a Caribbean Sclerosis? Stagnating Economic Growth in the Caribbean. Washington, DC: Inter-American Development Bank.

    • Search Google Scholar
    • Export Citation

    Thacker, Nita, SebastianAcevedo, RobertoPerrelli, JoongShik Kang, and MelesseTashu. 2013. “Economic Growth.” In The Eastern Caribbean Economic and Currency Union: Macroeconomics and Financial Systems, edited by A.Schipke, A.Cebotari, and N.Thacker,79–94. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation
1

The analysis in this book focuses on 13 Caribbean economies, divided into two analytical groups: nine tourism-intensive economies (Antigua and Barbuda, The Bahamas, Barbados, Dominica, Grenada, Jamaica, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines) and four commodity exporters (Belize, Guyana, Suriname, Trinidad and Tobago). Following IMF (2016b), small states are defined as those with populations of less than 1.5 million that are not advanced market economies or high-income oil-exporting countries. Annex 2.1 provides further details on the data used in the analysis.

2

The prediction is based on the following equation, estimated for 148 advanced and emerging market and developing economies on the 2000–15 sample: git = αt + βyit-N + εit, where g is average per capita growth over each five-year period t (N = 5), y is initial (log) per capita income, and αt, is a time fixed effect capturing external (global) factors.

3

The analysis adjusts the usual approach for estimating TFP by accounting for the destructive effects of natural disasters on the capital stock (see Annex 2.2). Without this adjustment, the analysis would attribute disaster damage to TFP growth, resulting in even weaker TFP growth.

4

Mapping how each of these determinants influences the contributions of capital and labor inputs and TFP is not straightforward. Nevertheless, there are reasons to expect that the variables under consideration influence investment, labor supply and demand, and TFP growth.

5

Migration data for 182 countries during 1980–2010 yield the following estimation results:

ln(Hit/ Hit –30) = 38.4–17.6 ln(Hit – 30) – 33.1 {High-skill E}it – 30:t – 17.3 {Medium-skill E}it – 30:t – 0.1 {Low-skill E},

where H denotes human capital and E denotes emigration as a percentage of all country nationals. The measure of H comes from the Penn World Tables and is based on the average years of schooling (from Barro and Lee 2013; Cohen and Leker 2014) and an assumed rate of return to education based on Mincer equation estimates (Psacharopoulos 1994). The results imply that countries with a lower initial level of H had faster growth in H (convergence). They also imply that the departure of the 30 percent of nationals with tertiary education from the Caribbean over this 30-year period has reduced Caribbean H by about 10 percent.

6

The highest crime rates are found in The Bahamas, Jamaica, and Trinidad and Tobago.

7

The costs comprise lethal and nonlethal victimization and forgone income of prison populations, private spending on security by businesses and households, and public spending, including the costs to the justice system, spending on police services, and spending on prisons.

8

The Ease of Doing Business indicator used here is the “distance to frontier” overall index, which summarizes 41 indicators for 10 doing business topics.

9

Following IMF (2017a), trade connectivity is measured by five indicators, which are here combined into a single composite indicator (first principal component): the number of countries exported to, the number of countries imported from, centrality in the world trade network (“eigencentrality”), centrality among exporters in the world trade network (“out eigencentrality”), and centrality among importers in the world trade network (“in eigencentrality”). Caribbean economies’ limited number of trading partners partially reflects the importance of tourism, given that trade in this sector is naturally more concentrated with regions of geographic proximity (North America and Europe).

10

The equations estimated are as follows: git = αt + βXit – N + εit, where g is average per capita growth over each five-year period t (N = 5), X is a matrix of country characteristics, and αt is a time fixed effect.

11

Cooks distance measures how influential observations are on the estimated regression coefficients (Cook 1979; Hayashi 2000). Observations with Cook’s distance greater than 4/N, where N is the sample size, are discarded.

12

A rise in the homicide rate by 10 people in 100,000 (approximately equivalent to moving from the 25th to the 75th percentile of the sample) is associated with about 0.3 percentage point lower average growth over the subsequent five years in all the specifications estimated. The result is both statistically and economically significant, and is consistent with the findings of Demirci, Moreno, and Wong (forthcoming), who use an instrumental variables approach based on deportations from the United States.

13

Costa Rica provides an example of significant gains in medium-term growth following reforms. The shift to faster growth in the 1990s was largely driven by Intel Corporation’s decision to place its manufacturing plant in the country, which, in turn, was motivated by Costa Rica’s high levels of educational attainment; economic openness; stable political, social, and macroeconomic environment; and strong doing business climate, reflecting decades of reforms, including investment in education (IMF 2016a).

    Other Resources Citing This Publication